MultiSlav: Using Cross-Lingual Knowledge Transfer to Combat the Curse of Multilinguality
This addresses the under-studied problem of Slavic language translation for hundreds of millions of speakers, though it appears incremental in extending existing multilingual approaches.
The paper tackles the problem of whether multilingual neural machine translation causes performance degradation or enables cross-lingual knowledge transfer, particularly for Slavic languages, and demonstrates cross-lingual benefits including in zero-shot translation for low-resource languages while providing state-of-the-art open-source models.
Does multilingual Neural Machine Translation (NMT) lead to The Curse of the Multlinguality or provides the Cross-lingual Knowledge Transfer within a language family? In this study, we explore multiple approaches for extending the available data-regime in NMT and we prove cross-lingual benefits even in 0-shot translation regime for low-resource languages. With this paper, we provide state-of-the-art open-source NMT models for translating between selected Slavic languages. We released our models on the HuggingFace Hub (https://hf.co/collections/allegro/multislav-6793d6b6419e5963e759a683) under the CC BY 4.0 license. Slavic language family comprises morphologically rich Central and Eastern European languages. Although counting hundreds of millions of native speakers, Slavic Neural Machine Translation is under-studied in our opinion. Recently, most NMT research focuses either on: high-resource languages like English, Spanish, and German - in WMT23 General Translation Task 7 out of 8 task directions are from or to English; massively multilingual models covering multiple language groups; or evaluation techniques.